Hip fractures are common among older women and can have a devastating impact on their ability to remain independent. A clinically important functional decline and failure to recover following a hip fracture has been documented as much as a year after the fracture, even among individuals who were functioning at high levels before the event. Age-associated androgen deficiency in women contributes to deficits in muscle mass, strength and power that are common in this patient population before the fracture, and are exacerbated afterward. A pilot study of testosterone (T) supplementation in elderly female hip fracture patients has demonstrated the feasibility of T treatment in this population, and showed gains in lean body mass (LBM) and muscle strength with active drug, compared to placebo. The benefits of exercise in restoring muscle strength and physical function after a hip fracture have been documented. However, it remains unclear whether T treatment can augment the effects of exercise on mobility and patient-reported function, or whether any observed benefits are sustained beyond the period of active treatment. Proposed is a 3-group, multi-center, randomized, placebo-controlled, double-blinded, parallel group clinical trial in frail elderly female hip fracture patients. 300 female hip fracture patients will be enrolled from 6 clinical sites, using objective screening criteria for T deficiency (serum total testosterone level < 30 ng/dl) and physical frailty (Modified Physical Performance Test (PPT) Score < 28). The trial will compare the effects of supervised exercise training (EX) alone, EX combined with T therapy (EX+T) and no EX with placebo T treatment (CON), to ascertain the incremental impact of adding T to ET in older adult women following hip fracture. The 6-month intervention will be followed by a 6-month no-treatment sustainability phase. The primary outcome measure is the Six Minute Walk Distance (6MWD). Secondary outcome measures include: 1) dual energy x-ray absorptiometry (DXA) measurements of whole body and appendicular LBM and bone mineral density of the unfractured proximal femur; 2) maximal skeletal muscle strength (1-RM) for leg extension in both limbs; 3) objective physical performance measures; and 4) self-reported performance of activities of daily living and quality of life, including the Hip Rating Questionnaire (HRQ). We plan to carefully monitor testosterone levels, adverse events, biochemical parameters, and factors related to adherence to the interventions. Information from this study has the potential to alter treatment of hip fracture in older women, a problem that contributes to significant morbidity and mortality, and has a large public health impact. The proposed study is highly aligned with NIA?s mission of identifying interventions that target common geriatric conditions, and improve treatment options for older adults with multiple morbidities or risk factors.